{"id":25,"date":"2026-05-31T18:46:12","date_gmt":"2026-05-31T18:46:12","guid":{"rendered":"https:\/\/onyxdigi.com\/blog\/?p=25"},"modified":"2026-05-31T20:26:53","modified_gmt":"2026-05-31T20:26:53","slug":"11-of-12-ai-wireframes-were-bad-heres-why","status":"publish","type":"post","link":"https:\/\/onyxdigi.com\/blog\/2026\/05\/31\/11-of-12-ai-wireframes-were-bad-heres-why\/","title":{"rendered":"11 of 12 AI Wireframes Were Bad \u2014 Here’s Why"},"content":{"rendered":"\n

This morning I generated 12 wireframe variants for a UAE wealth-management app. Start to finish: 47 minutes.<\/p>\n\n\n\n

Here’s the part nobody puts in the AI demos: 11 of them were bad.<\/p>\n\n\n\n

Not technically<\/em> bad \u2014 that’s the trap. The shapes were right. The hierarchy was readable. The spacing was clean. AI-generated wireframes almost always look correct on first glance, and that’s exactly what makes them dangerous. “Looks correct” and “is right” are not the same thing, and the distance between them is the entire job.<\/p>\n\n\n\n

Let me walk you through the eleven I threw away, because the reasons are the point.<\/p>\n\n\n\n

Four of them showed “interest earned.”<\/strong> This is a Sharia-compliant product. Interest \u2014 riba \u2014 isn’t a styling choice you swap out later; it’s forbidden, full stop. A wealth app for this market speaks in terms of profit-sharing and expected returns, not interest. Four variants were dead on arrival \u2014 and nothing about the layout told you that. You had to already know it.<\/p>\n\n\n\n

Three of them put the chart above the action.<\/strong> Looks fine \u2014 every fintech screenshot on Dribbble leads with a big chart. But this is a wealth app, and the person opening it didn’t come to admire a line going up. They came to do<\/em> something: top up, rebalance, move money toward a goal. Bury the action under a hero chart and you’ve optimised for looking like a finance app instead of working like one.<\/p>\n\n\n\n

Two of them used a US-style risk slider<\/strong> \u2014 Conservative on the left, Aggressive on the right. The mechanic is fine; the framing is foreign. For a Khaleeji investor, the mental model isn’t timid-versus-bold. It’s closer to Preservation \u2192 Growth: protecting what you’ve built while it grows. Same control, completely different language \u2014 and the wrong words quietly tell the user this product was designed for someone else.<\/p>\n\n\n\n

One of them was pink and purple.<\/strong> Wrong for this app, wrong for that market, wrong for any fintech anywhere. Some rejections take six seconds.<\/p>\n\n\n\n

And one of them was actually great.<\/strong><\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

That’s the real story. Not “AI made twelve wireframes” \u2014 twelve is easy. The story is that someone had to look at twelve confident, polished, correct-looking screens and know, fast and for specific reasons, that eleven of them would quietly hurt the product.<\/p>\n\n\n\n

And the twelfth one? It got right exactly what the others missed: profit-sharing language instead of interest, the primary action sitting above the chart, a Preservation \u2192 Growth frame the user would actually recognise, a restrained palette that looked like money rather than a candy wrapper. On the surface it wasn’t more designed<\/em> than the rejects \u2014 same boxes, same grid. It was simply correct for the brief. The difference between it and the other eleven was invisible to the eye and obvious to the judgement.<\/p>\n\n\n\n

The narrative everyone’s selling gets this backwards<\/h2>\n\n\n\n

“AI replaces designers” misses the entire mechanism of the work. Here’s what actually happened in my morning.<\/p>\n\n\n\n

AI got me 12 variants in 47 minutes. Two years ago that was two days.<\/mark><\/p>\n\n\n\n

Then the senior designer’s job kicked in: throw away 11 of them in 6 minutes<\/strong> \u2014 and be able to say exactly<\/em> why each one died. Riba. Chart-before-action. Foreign risk language. Off-brand colour. Not “I don’t like it.” Specific, defensible, rooted in the market and the constraints.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

That triage is<\/em> the work now. The generation got cheap. The judgement about what to keep didn’t move at all.<\/p>\n\n\n\n

What the 6 minutes is actually made of<\/h2>\n\n\n\n

It looks like 6 minutes. It’s really two decades.<\/em><\/strong><\/p>\n\n\n\n

To kill the “interest earned” variants in seconds, you have to already carry the knowledge that riba is non-negotiable here \u2014 you can’t research it after the fact, because the AI will hand you the haram version with total confidence and a beautiful layout. To know the chart belongs below the action, you have to understand why the user opened the app<\/em> \u2014 intent, not aesthetics. To rewrite Conservative\u2013Aggressive as Preservation\u2013Growth, you have to have sat with enough Gulf investors to feel how they actually think about their money.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

None of that is in the wireframe. All of it is in the person reviewing the wireframe. That gap \u2014 between a screen that looks right and a screen that is right for this<\/em> user, in this<\/em> market, under these<\/em> rules \u2014 is where experience lives. AI doesn’t close that gap. If anything it widens it, by producing wrong answers that look more convincing than ever.<\/p>\n\n\n\n

So where does that leave us<\/h2>\n\n\n\n

Faster, if you have judgement. Exposed, if you don’t.<\/p>\n\n\n\n

The designer who treats AI as a junior working at superhuman speed \u2014 generate a lot, then cut ruthlessly \u2014 is now operating at a level that was impossible eighteen months ago. The designer who trusts the output because it “looks correct” is shipping the riba screen to a Sharia-compliant bank and finding out the hard way.<\/p>\n\n\n\n

Speed is a commodity now. Anyone can generate twelve wireframes before their coffee’s cold. What can’t be commoditised is knowing which eleven to throw away, and why.<\/p>\n\n\n\n

That triage is the work. That’s what 23 years buys you.<\/p>\n\n\n\n

Speed is a commodity. Judgement is not.<\/p>\n","protected":false},"excerpt":{"rendered":"

This morning I generated 12 wireframe variants for a UAE wealth-management app. Start to finish: 47 minutes. Here’s the part nobody puts in the AI demos: 11 of\u2026<\/p>\n","protected":false},"author":1,"featured_media":30,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3,4,5],"tags":[10,15,19,12],"class_list":["post-25","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ux","category-arabic-ux","category-fintech-ux","tag-ai","tag-fintech","tag-productdesign","tag-ux"],"_links":{"self":[{"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/posts\/25","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/comments?post=25"}],"version-history":[{"count":2,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/posts\/25\/revisions"}],"predecessor-version":[{"id":33,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/posts\/25\/revisions\/33"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/media\/30"}],"wp:attachment":[{"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/media?parent=25"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/categories?post=25"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/onyxdigi.com\/blog\/wp-json\/wp\/v2\/tags?post=25"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}